How a 77 outranks an 88 in the R score

In the worked example the Bureau de coopération interuniversitaire uses to explain the R score, Vincent finishes a course with a 77 and Olivier, in another class, with an 88. Vincent’s R score is 32.91 and Olivier’s is 32.39. The R score does not rank grades. It ranks where a student stands in the group that took the course, then corrects for how strong and how spread out that group was.

The formula, read at source

Quebec’s universities adopted the R score, the cote de rendement au collégial, in 1995, and BCI publishes its rules. Its formula document (21 January 2025) prints it as R = ((Z × IDGZ) + IFGZ + 5) × 5. Z is the college Z score, (grade − average) ÷ standard deviation, worked over the evaluation group: the students who took the same course at the same college in the same term and were graded the same way. Grades below 50 are left out of the average and the standard deviation, and the standard deviation divides by the number of grades. IFGZ and IDGZ are the average and the standard deviation of the same students’ secondary-school Z scores, from the Secondary 4 and 5 subjects with ministry exams. BCI caps IFGZ at ±2, IDGZ between 0.5 and 1.5 and Z × IDGZ at ±3, so every R score lands between 0 and 50; most, it says, between 15 and 35.

Three classes, worked

BCI’s example has three fictitious classes of nine. Class A scored 81 to 89 (average 85), class B 71 to 79 (average 75) and class C 59 to 91 (average 75). Pick six students by grade and you take class A’s top four and class C’s top two; class B gets no one. Turn the grades into Z scores and classes A and B become identical, 1.55 at the top down to −1.55, because both spread their grades the same way (a standard deviation of 2.58). Class C spreads them four times as wide (10.33), and its Z scores match too: the top two of each class now tie.

Then BCI assumes the indicators: class A weak (IFGZ 0.55, IDGZ 0.80), class B strong (1.12 and 0.60), class C mixed (0.72 and 1.00). Camille, second in class B with a 78, works out as (1.16 × 0.60 + 1.12 + 5) × 5 = 34.08. The six best are now class A’s first, class B’s top three and class C’s top two:

Vincent’s 77 takes the sixth place; Catherine’s 83 and Olivier’s 88 come next. The same thing happens at the bottom: Audrey, last in class B with a 71, scores 25.95, above three class A students with 81 to 83.

Two decimals in BCI’s table

All 27 of BCI’s R scores reproduce exactly once Z is rounded to two decimals, as its table prints it. Kept at full precision, 18 of the 27 would print differently: Camille’s Z is 1.162 and her R score 34.086, the largest gap is 0.023, at Catherine and Jacob, and the order of the 27 does not change. BCI’s line for Camille shows her corrected Z as 1.82, but (1.82 + 5) × 5 is 34.10: its 34.08 comes from the unrounded 1.816. The calculator keeps every digit.

The average universities read

The average R score weights each course by its units: BCI’s document multiplies the R score of a 2.66-unit course by 2.66. Since the winter 2005 admissions, failures count less: a course failed in the first term at college counts a quarter of its units, a course failed later counts half. Take the example the calculator opens with, four passed courses whose R scores of 27.8 to 31.15 average 29.657. A first-term failure at 16.2 over 2.66 units brings the average to 28.760; the same failure in a later term would leave it at 27.976, and at full weight it would be 26.669.

Which average a university reads is set by rules the calculator does not apply: BCI’s documents point to the program that leads to the diploma, or to the last program if at least 16 courses count toward it, with the overall average as the fallback, and McGill says it reviews CEGEP applicants on the overall R score.

What the transcript will not tell you

The group average on a CEGEP transcript counts every grade, while BCI’s leaves out the grades below 50, which is why, its questions and answers explain, a student well above the transcript average can get a lower R score than expected. The Cégep de Chicoutimi’s guide adds that the evaluation group’s average does not appear on the transcript at all, so a student cannot compute their own R score exactly. The two indicators come from the ministry, which recalculates the R scores of the last six terms four times a year, so a course’s R score can move after the fact; BCI’s own example is a 23 that became a 22.

Two more rules matter at the edges. In a high, tight group, BCI’s ceiling rule stretches the grades above the average so that a 100 reaches a weighted Z of 2 (an R score of 35 when IFGZ is negative): in a group averaging 92 with a standard deviation of 5, a 96 scores 30.000 instead of 29.000. And one case is held here rather than resolved, as of October 9, 2026. For a group with fewer than six students who have Quebec secondary results and a grade of 50 or more, the questions-and-answers document says no R score is calculated, while the formula document sets IFGZ to 0 and IDGZ to 1 and calculates one without correction. The calculator’s defaults are those two values, and it does not decide which document governs your group.